Johns Hopkins University, 6 Health Resources and Services Administration, 7

Transcription

1 Michael Mugavero, 1 Andrew Westfall, 1 Anne Zinski, 1 Jessica Davila, 2 Mari-Lynn Drainoni, 3 Lytt Gardner, 4 Jeanne Keruly, 5 Faye Malitz, 6 Gary Marks, 4 Lisa Metsch, 7 Tracey Wilson, 8 Thomas Giordano, 2 for the Retention in Care (RIC) Study Group* 1 University of Alabama at Birmingham, 2 Baylor College of Medicine, 3 Boston University Medical Center, 4 Centers for Disease Control and Prevention, 5 Johns Hopkins University, 6 Health Resources and Services Administration, 7 University of Miami, 8 State University of New York, Downstate Medical Center The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention

2 Background Measuring retention in care is complex Multiple visits at varying intervals over time Numerous retention measures employed Missed visit ( no show ) & kept visit measures Each associated w/ biological & clinical outcomes Most studies focus on a single measure Degree to which measures are related to one another and to outcomes largely unexplored

3 Missed Visits Appt. Adherence Mugavero, Davila, Nevin & Giordano. AIDS Pt Care STDs 2010;24 Visit Constancy Gap in Care HRSA HAB Measure Patient A Yes; 1 80% 100% No Yes Patient B Yes; 4 33% 50% Yes Yes Patient C No; 0 100% 75% No Yes Patient D Yes; 1 67% 25% Yes No

4 Methods Study aims: Evaluate correlation among 6 retention measures Evaluate prognostic value of measures in predicting plasma viral load (VL) suppression Design: Retention in Care (RIC) Intervention Study Six academically-affiliated HIV clinics Phase I (Clinic-wide) & Phase II (Behavioral RCT) Current study: Clinic-wide cohort design during 12 months preceding Phase I RIC intervention

5 Methods Study period: May 2008 April 2009 Eligibility criteria: Attended 1 primary HIV care appointment in the year preceding study period 1 scheduled primary HIV care appointment during 1 st six months of study period Criteria employed to identify established clinic patients in whom retention could be measured

6 Methods Principal outcome: VL suppression (<400 c/ml) at end of study period 12-month VL suppression VL nearest 30 April 2009 (+ 120 days) Principal exposures: Six commonly used retention measures Scheduled visits w/ primary HIV medical provider Calculated based upon kept and no show visits

7 Methods Statistical analyses: Spearman rank correlation: compare measures Logistic regression for each measure (VL<400 c/ml) C-statistic: discriminatory capacity of measures Prognostic value to assign pts to correct 12-mo VL state Range ( coin toss perfect discrimination) Estimate of area under ROC curve Primary analyses: pts w/ missing 12-mo VL excluded Sensitivity analyses: pts w/ missing 12-mo VL=failure

8 Methods: Retention measures Measure Missed visits: count Missed visits: dichotomous Visit adherence Description Number of no show visits accrued (count measure) 1 no show visit (dichotomous measure, no = retained) Proportion of kept visits / (kept + no-show visits) (continuous measure, range= )

9 Methods: Retention measures Measure Missed visits: count Missed visits: dichotomous Visit adherence 4-month constancy 6-month gap HRSA HAB Description Number of no show visits accrued (count measure) 1 no show visit (dichotomous measure, no = retained) Proportion of kept visits / (kept + no-show visits) (continuous measure, range= ) Number of 4-month intervals with at least 1 kept visit (categorical measure, range=0-3) 189 days elapsed between sequential kept visits (dichotomous measure, no = retained) 2 kept visits separated by 90 days (dichotomous measure, yes = retained)

10

11 Baseline characteristics (n=10,053) Data presented as mean + standard deviation or n (%)

12 Baseline characteristics (n=10,053) Data presented as mean + standard deviation or n (%)

13 Data presented as mean + standard deviation or n (%)

14 Spearman rank correlation matrix

15 Association of retention measures with 12-month VL suppression (<400 c/ml) a OR presented per missed visit (count), per 0.5 increase for visit adherence & 4-month constancy, and retained for dichotomous retention measures: missed visits, 6-month gap, and HRSA HAB measure b Sn and Sp for cut-points for retained of: 1 missed visits (count), 70% visit adherence, and attended visits in all 3 intervals for 4-month constancy, and per retained for dichotomous measures

16 (c-statistic=0.67)

17 (c-statistic=0.62)

18 (c-statistic=0.69)

19 (c-statistic=0.63)

20 (c-statistic=0.61)

21 (c-statistic=0.59)

22 Sensitivity analyses (M=F) 12-month VL suppression (<400 c/ml) a OR presented per missed visit (count), per 0.5 increase for visit adherence & 4-month constancy, and retained for dichotomous retention measures: missed visits, 6-month gap, and HRSA HAB measure

23 Conclusions Considerable variability among six measures in categorizing retention Wide range of correlation across measures Missed visit measures highly correlated ( ) Kept visit measures highly correlated ( ) Correlation lower across these two groups ( ) All six measures had significant association (P<0.001) with 12-month VL suppression

24 Limitations Observational study: cannot ascribe causality Exclusion of patients new to care ART exposure not systematically captured during 12-month study period Relatively short observation period Modest discriminatory capacity of measures Augmented by ART receipt & adherence?

25

26 Implications No clear gold standard to measure retention Each measure may have value and utility according to setting and circumstance Merit in using a missed and a kept visit based measure in research settings Measures capturing different constructs? Missed visit based Adherence Kept visit based Persistence

27 Acknowledgments Boston University Medical Center Mari-Lynn Drainoni (PI) Cintia Ferreira Lisa Koppelman Maya McDoom Michal Naisteter Karina Osella Glory Ruiz Paul Skolnik Meg Sullivan (PI) SUNY Downstate Medical Center Sophia Gibbs-Cohen Elana Desrivieres Mayange Frederick Kevin Gravesande Susan Holman Harry Johnson Tonya Taylor Tracey Wilson (PI) University of Alabama-Birmingham Scott Batey Stephanie Gaskin Michael Mugavero (PI) Jill Murphree Jim Raper Michael Saag (PI) Suneetha Thogaripally James Willig Anne Zinski Baylor College of Medicine Monisha Arya David Bartholomew Tawanna Biggs Hina Budhwani Jessica Davila Christine Jacobsen Tom Giordano (PI) Nancy Miertschin Shapelle Payne William Slaughter Johns Hopkins University Mollie Jenckes Jeanne Keruly (PI) Angie McCray Mary McGann Richard Moore (PI) Melissa Otterbein LiMing Zhou University of Miami Carolyn Garzon Jesline Jean-Simon Kathy Mercogliano Lisa Metsch (PI) Allan Rodriguez (PI) Gilbert Saint-Jean Marvin Shika Mountain Plains AETC Lucy Bradley-Springer Marla Corwin Federal Laura Cheever, HRSA Faye Malitz, HRSA Robert Mills, HRSA Jason Craw, CDC/ICF Lytt Gardner, CDC Sonali Girde, CDC/ICF Gary Marks, CDC We thank the study participants, providers, clinical and research personnel at the six study sites as well as the CDC and HRSA administrative and data management teams. Supported by CDC & HRSA via CDC contracts: , , , , , MJM supported by 5K23MH

28 Questions?